https://aiweekly.co/alerts/singh-detecting-ai-agent-failures-isnt-enough-to-govern AI Weekly en 2026-09-16T21:12:02+00:00 Singh: Detecting AI Agent Failures Isn't Enough to Govern https://aiweekly.co/alerts/tu-dresden-on-premise-ai-nears-cloud-accuracy-in-nature-paper AI Weekly en 2026-09-16T21:12:02+00:00 TU Dresden on-premise AI nears cloud accuracy in Nature paper https://aiweekly.co/alerts/jd-vance-rejects-ai-rules-tells-labs-to-stop-frankenstein AI Weekly en 2026-09-16T21:12:02+00:00 JD Vance rejects AI rules, tells labs to stop 'Frankenstein' https://aiweekly.co/alerts/google-home-mcp-lets-claude-chatgpt-control-nest-devices AI Weekly en 2026-09-16T19:55:59+00:00 Google Home MCP Lets Claude, ChatGPT Control Nest Devices https://aiweekly.co/alerts/may-mobility-to-list-on-nasdaq-via-14b-acp-spac-deal AI Weekly en 2026-09-16T19:55:59+00:00 May Mobility to List on Nasdaq via $1.4B ACP SPAC Deal https://aiweekly.co/alerts/biren-mulls-1b-share-placement-third-raise-since-hk-ipo AI Weekly en 2026-09-16T18:32:03+00:00 Biren Mulls $1B Share Placement, Third Raise Since HK IPO https://aiweekly.co/alerts/anthropic-folds-claude-cowork-into-main-claude-chat AI Weekly en 2026-09-16T18:12:01+00:00 Anthropic folds Claude Cowork into main Claude chat https://aiweekly.co/alerts/ai-scribe-invents-drug-use-in-australian-patients-record AI Weekly en 2026-09-16T18:12:01+00:00 AI scribe invents drug use in Australian patient's record https://aiweekly.co/alerts/von-der-leyen-europe-neednt-build-frontier-ai-itself AI Weekly en 2026-09-16T18:12:01+00:00 Von der Leyen: Europe needn't build frontier AI itself https://aiweekly.co/alerts/barrett-urges-skepticism-of-doomer-ceos-and-the-china-canard AI Weekly en 2026-09-16T18:12:01+00:00 Barrett urges skepticism of doomer CEOs and the China canard https://aiweekly.co/alerts/openais-project-lily-pays-contractors-to-read-chatgpt-chats AI Weekly en 2026-09-16T18:12:01+00:00 OpenAI's 'Project Lily' Pays Contractors to Read ChatGPT Chats https://aiweekly.co/alerts/tu-dresdens-on-premise-medical-ai-hits-90-on-mimic-iv AI Weekly en 2026-09-16T18:12:01+00:00 TU Dresden's on-premise medical AI hits 90% on MIMIC-IV https://aiweekly.co/alerts/on-prem-qwen-35-matches-cloud-gpt-52-on-nature-med-tests AI Weekly en 2026-09-16T18:12:01+00:00 On-prem Qwen-3.5 matches cloud GPT-5.2 on Nature Med tests https://aiweekly.co/alerts/fda-clears-powerful-medicals-queen-of-hearts-ekg-ai-for-stemi AI Weekly en 2026-09-16T18:12:01+00:00 FDA clears Powerful Medical's Queen of Hearts EKG AI for STEMI https://aiweekly.co/alerts/paper-locates-and-steers-llama-31-opportunity-recognition-dial AI Weekly en 2026-09-16T18:12:01+00:00 Paper locates and steers Llama 3.1 'opportunity recognition' dial https://aiweekly.co/alerts/deepmind-opens-institute-to-publish-essays-on-agi-safety AI Weekly en 2026-09-16T18:12:01+00:00 DeepMind opens Institute to publish essays on AGI safety https://aiweekly.co/alerts/anthropic-unifies-claude-chat-and-cowork-adds-slide-export AI Weekly en 2026-09-16T16:57:24+00:00 Anthropic Unifies Claude Chat and Cowork, Adds Slide Export https://aiweekly.co/alerts/cohere-signs-definitive-merger-with-aleph-alpha-at-20b-valuation AI Weekly en 2026-09-16T16:57:24+00:00 Cohere Signs Definitive Merger With Aleph Alpha at ~$20B Valuation https://aiweekly.co/alerts/legg-opens-deepmind-institute-calls-agi-claims-premature AI Weekly en 2026-09-16T16:57:24+00:00 Legg Opens DeepMind Institute, Calls AGI Claims Premature https://aiweekly.co/alerts/pentagon-cto-michael-rejects-us-ai-stakes-new-ai-rules AI Weekly en 2026-09-16T16:57:24+00:00 Pentagon CTO Michael Rejects US AI Stakes, New AI Rules https://aiweekly.co/alerts/stepaudio-3-realtime-posts-989-on-full-duplex-voice-bench AI Weekly en 2026-09-16T15:53:46+00:00 StepAudio 3 Realtime posts 98.9 on Full-Duplex voice bench https://aiweekly.co/alerts/stepaudio-3-music-drafts-abc-notation-before-530-tracks AI Weekly en 2026-09-16T15:53:46+00:00 StepAudio 3 Music drafts ABC notation before 5:30 tracks https://aiweekly.co/alerts/chinese-led-paper-maps-five-stage-path-to-self-improving-ai AI Weekly en 2026-09-16T15:53:46+00:00 Chinese-led paper maps five-stage path to self-improving AI https://aiweekly.co/alerts/openai-pilots-sponsored-agents-in-chatgpt-with-wayfair-angi AI Weekly en 2026-09-16T15:28:03+00:00 OpenAI Pilots Sponsored Agents in ChatGPT With Wayfair, Angi https://aiweekly.co/alerts/openai-anthropic-staff-blindsided-by-pace-the-frontier-plan AI Weekly en 2026-09-16T13:57:33+00:00 OpenAI, Anthropic Staff Blindsided by 'Pace the Frontier' Plan https://aiweekly.co/alerts/openai-pushes-reverse-federalism-as-thune-ai-bill-looms AI Weekly en 2026-09-16T13:57:33+00:00 OpenAI Pushes 'Reverse Federalism' as Thune AI Bill Looms https://aiweekly.co/alerts/controlai-runs-hundreds-of-hill-meetings-on-ai-extinction-risk AI Weekly en 2026-09-16T13:57:33+00:00 ControlAI Runs Hundreds of Hill Meetings on AI Extinction Risk https://aiweekly.co/alerts/meta-one-bundles-muse-ai-across-facebook-instagram-whatsapp AI Weekly en 2026-09-16T12:31:25+00:00 Meta One Bundles Muse AI Across Facebook, Instagram, WhatsApp https://aiweekly.co/alerts/hackers-rip-open-a-flock-camera-50200-vehicles-in-21-days AI Weekly en 2026-09-16T12:31:25+00:00 Hackers rip open a Flock camera; 50,200 vehicles in 21 days https://aiweekly.co/alerts/zron-documents-show-ai-packaging-stolen-files-as-police-intel AI Weekly en 2026-09-16T12:31:25+00:00 ZRON documents show AI packaging stolen files as police intel https://aiweekly.co/alerts/hutchins-agentic-ai-cyberattack-panic-is-marketing-bs AI Weekly en 2026-09-16T12:12:02+00:00 Hutchins: 'Agentic AI' Cyberattack Panic Is Marketing BS https://aiweekly.co/alerts/anthropic-signs-32b-queensland-data-centre-deal-for-claude AI Weekly en 2026-09-16T10:57:07+00:00 Anthropic Signs $32B Queensland Data Centre Deal for Claude https://aiweekly.co/alerts/agibots-harnessvln-posts-76-on-hm3d-v2-training-free AI Weekly en 2026-09-16T09:25:46+00:00 AGIBOT's HarnessVLN posts 76% on HM3D-v2 training-free https://aiweekly.co/alerts/ddo-weight-edit-defense-cuts-llm-abliteration-asr-below-10 AI Weekly en 2026-09-16T09:25:46+00:00 DDO Weight-Edit Defense Cuts LLM Abliteration ASR Below 10% https://aiweekly.co/alerts/six-nation-survey-citizens-want-ai-rules-not-deregulation AI Weekly en 2026-09-16T09:12:01+00:00 Six-nation survey: citizens want AI rules, not deregulation https://aiweekly.co/alerts/boxs-levie-says-diffusion-not-intelligence-gates-enterprise-ai AI Weekly en 2026-09-16T09:12:01+00:00 Box's Levie says diffusion, not intelligence, gates enterprise AI https://aiweekly.co/alerts/boxs-aaron-levie-joins-sequoia-podcast-on-ai-era-reinvention AI Weekly en 2026-09-16T09:12:01+00:00 Box's Aaron Levie joins Sequoia podcast on AI-era reinvention https://aiweekly.co/alerts/deepseek-engineer-compares-anthropic-agi-lead-to-hitlers-bomb AI Weekly en 2026-09-16T06:30:32+00:00 DeepSeek Engineer Compares Anthropic AGI Lead to Hitler's Bomb https://aiweekly.co/alerts/sk-hynix-in-talks-with-intel-on-first-us-memory-chip-production AI Weekly en 2026-09-16T06:30:32+00:00 SK Hynix in Talks With Intel on First US Memory Chip Production https://aiweekly.co/alerts/emergence-ai-stress-tests-long-horizon-multi-agent-safety AI Weekly en 2026-09-16T06:30:32+00:00 Emergence AI Stress-Tests Long-Horizon Multi-Agent Safety https://aiweekly.co/alerts/janelia-google-male-fly-connectome-goes-viral-on-minecraft-demo AI Weekly en 2026-09-16T04:54:33+00:00 Janelia-Google Male Fly Connectome Goes Viral on Minecraft Demo https://aiweekly.co/alerts/nus-ntu-survey-maps-foundation-models-across-game-lifecycle AI Weekly en 2026-09-16T04:54:33+00:00 NUS, NTU Survey Maps Foundation Models Across Game Lifecycle https://aiweekly.co/alerts/sciencebuddy-paper-nests-harness-evolution-inside-rl-loop AI Weekly en 2026-09-16T04:54:33+00:00 ScienceBuddy paper nests harness evolution inside RL loop https://aiweekly.co/alerts/jhu-paper-lifts-llm-long-horizon-retention-from-12-to-349 AI Weekly en 2026-09-16T03:54:28+00:00 JHU paper lifts LLM long-horizon retention from 1.2% to 34.9% https://aiweekly.co/alerts/sciencebuddy-paper-ships-recursive-in-recursive-agent-workspace AI Weekly en 2026-09-16T03:54:28+00:00 ScienceBuddy paper ships recursive-in-recursive agent workspace https://aiweekly.co/alerts/cyber-veterans-say-altman-amodei-shut-them-out-of-safety-plans AI Weekly en 2026-09-16T03:29:31+00:00 Cyber Veterans Say Altman, Amodei Shut Them Out of Safety Plans https://aiweekly.co/alerts/princetons-meredith-martin-previews-mlas-ai-framework AI Weekly en 2026-09-16T03:12:01+00:00 Princeton's Meredith Martin previews MLA's AI framework https://aiweekly.co/alerts/semianalysis-just-23-gw-of-us-datacenters-delayed-by-moratoriums AI Weekly en 2026-09-16T01:55:18+00:00 SemiAnalysis: Just 2.3 GW of US Datacenters Delayed by Moratoriums https://aiweekly.co/alerts/google-splits-design-from-code-in-vibe-design-agents-paper AI Weekly en 2026-09-16T01:55:18+00:00 Google Splits Design From Code in 'Vibe Design Agents' Paper https://aiweekly.co/alerts/sails-poison-set-choice-swings-llama-3-8b-backdoors-3-to-80 AI Weekly en 2026-09-16T00:25:33+00:00 SAILS: Poison-Set Choice Swings LLaMA-3-8B Backdoors 3% to 80% https://aiweekly.co/alerts/ilands-70000-ai-agents-flood-inboxes-with-cold-pitches AI Weekly en 2026-09-16T00:12:01+00:00 iLands' 70,000 AI agents flood inboxes with cold pitches https://aiweekly.co/alerts/developers-route-claude-code-at-rival-models-via-openrouter AI Weekly en 2026-09-15T21:27:46+00:00 Developers Route Claude Code at Rival Models via OpenRouter https://aiweekly.co/alerts/ngu-sampling-method-targets-rls-matthew-effect-in-llms AI Weekly en 2026-09-15T21:12:01+00:00 NGU sampling method targets RL's 'Matthew Effect' in LLMs https://aiweekly.co/alerts/canadian-teachers-analyst-label-k-12-ai-high-risk-now AI Weekly en 2026-09-15T21:12:01+00:00 Canadian teachers' analyst: label K-12 AI 'high-risk' now https://aiweekly.co/alerts/anthropic-researcher-jacob-coxon-quits-urges-ai-slowdown AI Weekly en 2026-09-15T21:12:01+00:00 Anthropic researcher Jacob Coxon quits, urges AI slowdown https://aiweekly.co/alerts/hugging-faces-delangue-demands-100m-compute-traces-from-openai AI Weekly en 2026-09-15T19:56:01+00:00 Hugging Face's Delangue Demands $100M Compute, Traces From OpenAI https://aiweekly.co/alerts/paper-orthrus-lossless-decoding-only-holds-under-fp32 AI Weekly en 2026-09-15T18:31:50+00:00 Paper: Orthrus 'Lossless' Decoding Only Holds Under FP32 https://aiweekly.co/alerts/lennard-ai-ceos-doom-talk-distracts-from-real-harms AI Weekly en 2026-09-15T18:12:01+00:00 Lennard: AI CEOs' Doom Talk Distracts From Real Harms https://aiweekly.co/alerts/libvips-bans-generative-ai-contributions-over-copyright-risk AI Weekly en 2026-09-15T18:12:01+00:00 libvips bans generative-AI contributions over copyright risk https://aiweekly.co/alerts/cms-launched-medicare-wiser-ai-pilot-despite-vendor-warning AI Weekly en 2026-09-15T18:12:01+00:00 CMS launched Medicare WISeR AI pilot despite vendor warning https://aiweekly.co/alerts/openai-confirms-weeks-of-safety-talks-with-anthropic-google AI Weekly en 2026-09-15T18:12:01+00:00 OpenAI confirms weeks of safety talks with Anthropic, Google https://aiweekly.co/alerts/salesforce-and-anthropic-put-37-sales-skills-into-claude AI Weekly en 2026-09-15T18:12:01+00:00 Salesforce and Anthropic put 37 sales skills into Claude https://aiweekly.co/alerts/kalshi-pulled-ai-compute-curve-after-commerce-order-semafor AI Weekly en 2026-09-15T16:57:44+00:00 Kalshi Pulled AI-Compute Curve After Commerce Order: Semafor https://aiweekly.co/alerts/musk-urges-us-chinese-labs-to-cross-test-ai-before-release AI Weekly en 2026-09-15T16:57:44+00:00 Musk Urges US, Chinese Labs to Cross-Test AI Before Release https://aiweekly.co/alerts/medgemma-27b-rarely-notices-swapped-x-rays-if-report-is-present AI Weekly en 2026-09-15T15:53:09+00:00 MedGemma-27B rarely notices swapped X-rays if report is present https://aiweekly.co/alerts/orthrus-lossless-decoding-holds-only-under-fp32-paper-finds AI Weekly en 2026-09-15T15:53:09+00:00 Orthrus 'lossless' decoding holds only under FP32, paper finds https://aiweekly.co/alerts/grouped-value-attention-cuts-kv-cache-45-47-vs-gqa-at-350m AI Weekly en 2026-09-15T15:53:09+00:00 Grouped Value Attention cuts KV cache 45-47% vs GQA at 350M https://aiweekly.co/alerts/salesforce-unveils-koa-a-crm-reasoning-model-on-nvidia-nemotron AI Weekly en 2026-09-15T15:27:03+00:00 Salesforce Unveils Koa, a CRM Reasoning Model on Nvidia Nemotron https://aiweekly.co/alerts/anthropic-co-founder-backs-verifiable-ai-kill-switch-rules AI Weekly en 2026-09-15T15:27:03+00:00 Anthropic Co-Founder Backs Verifiable AI Kill Switch Rules https://aiweekly.co/alerts/salvaggio-ai-rules-must-resist-radical-intentionalists AI Weekly en 2026-09-15T15:12:02+00:00 Salvaggio: AI rules must resist 'radical intentionalists' https://aiweekly.co/alerts/coester-koutsoupias-zbysiski-prove-k-server-conjecture AI Weekly en 2026-09-15T15:12:02+00:00 Coester, Koutsoupias, Zbysiński prove k-server conjecture https://aiweekly.co/alerts/singla-preprint-claims-matroid-secretary-conjecture-resolved AI Weekly en 2026-09-15T15:12:02+00:00 Singla preprint claims matroid secretary conjecture resolved https://aiweekly.co/alerts/song-zhang-prove-matrix-spencer-conjecture-with-8n-bound AI Weekly en 2026-09-15T15:12:02+00:00 Song, Zhang prove Matrix Spencer conjecture with 8√n bound https://aiweekly.co/alerts/lynnreal-omni-unifies-video-tasks-in-32b-dit-27b-flash-at-377ms AI Weekly en 2026-09-15T13:55:11+00:00 LynnReal-Omni Unifies Video Tasks in 32B DiT, 27B Flash at 377ms https://aiweekly.co/alerts/profound-hits-18b-on-180m-series-d-from-sequoia-kleiner AI Weekly en 2026-09-15T12:35:24+00:00 Profound Hits $1.8B on $180M Series D From Sequoia, Kleiner https://aiweekly.co/alerts/decree-841-takes-effect-lets-beijing-withhold-exit-ban-notices AI Weekly en 2026-09-15T12:35:24+00:00 Decree 841 Takes Effect, Lets Beijing Withhold Exit-Ban Notices https://aiweekly.co/alerts/microsoft-ships-emergency-windows-fix-for-rds-hyper-v-usb-audio AI Weekly en 2026-09-15T12:35:24+00:00 Microsoft Ships Emergency Windows Fix for RDS, Hyper-V, USB Audio https://aiweekly.co/alerts/ex-deepmind-safety-researcher-warns-ai-could-kill-us-all AI Weekly en 2026-09-15T10:57:05+00:00 Ex-DeepMind Safety Researcher Warns AI Could 'Kill Us All' https://aiweekly.co/alerts/apples-ios-27-adds-facetime-gore-blur-and-safari-ask-to-browse AI Weekly en 2026-09-15T10:57:05+00:00 Apple's iOS 27 Adds FaceTime Gore Blur and Safari Ask to Browse https://aiweekly.co/alerts/google-opens-claude-opus-5-to-all-engineers-via-antigravity AI Weekly en 2026-09-15T09:27:59+00:00 Google Opens Claude Opus 5 to All Engineers via Antigravity https://aiweekly.co/alerts/cunningham-trammell-paper-rsi-loops-not-yet-self-sustaining AI Weekly en 2026-09-15T09:27:59+00:00 Cunningham, Trammell Paper: RSI Loops Not Yet Self-Sustaining https://aiweekly.co/alerts/boston-dynamics-runs-robot-brains-on-tpus-rivals-stay-on-jetson AI Weekly en 2026-09-15T09:12:01+00:00 Boston Dynamics runs robot brains on TPUs; rivals stay on Jetson https://aiweekly.co/alerts/hazardauditor-claims-165-point-gain-over-prior-agent-guards AI Weekly en 2026-09-15T07:59:44+00:00 HazardAuditor Claims 16.5-Point Gain Over Prior Agent Guards https://aiweekly.co/alerts/kaininja-paper-extends-trellis2-3d-generator-to-part-level AI Weekly en 2026-09-15T07:59:44+00:00 KaiNinja Paper Extends TRELLIS.2 3D Generator to Part Level https://aiweekly.co/alerts/elo-per-token-study-parallel-sessions-beat-long-agent-runs AI Weekly en 2026-09-15T07:59:44+00:00 Elo-per-Token Study: Parallel Sessions Beat Long Agent Runs https://aiweekly.co/alerts/shanghai-ai-lab-ships-atria-dawn-preview-a-744b-agentic-moe AI Weekly en 2026-09-15T06:27:59+00:00 Shanghai AI Lab Ships Atria Dawn Preview, a 744B Agentic MoE https://aiweekly.co/alerts/rsiagent-training-free-memory-lifts-kimi-k3-glm-53-past-gpt-6 AI Weekly en 2026-09-15T06:27:59+00:00 RSIAgent: Training-Free Memory Lifts Kimi-K3, GLM-5.3 Past GPT-6 https://aiweekly.co/alerts/autopoietic-game-theory-links-cooperation-to-compute-cost AI Weekly en 2026-09-15T06:12:00+00:00 Autopoietic Game Theory links cooperation to compute cost https://aiweekly.co/alerts/ai-now-institute-europes-ai-market-is-structurally-captured AI Weekly en 2026-09-15T06:12:00+00:00 AI Now Institute: Europe's AI market is 'structurally captured' https://aiweekly.co/alerts/perplexity-ships-portable-computer-for-windows-on-24gb-rtx-gpus AI Weekly en 2026-09-15T05:00:05+00:00 Perplexity Ships Portable Computer for Windows on 24GB+ RTX GPUs https://aiweekly.co/alerts/sakana-ais-pc-alm-trains-1000-layer-nets-without-backprop AI Weekly en 2026-09-15T05:00:05+00:00 Sakana AI's PC-ALM trains 1,000-layer nets without backprop https://aiweekly.co/alerts/occamy-10-open-35b-edges-gpt-56-sol-on-claw-eval-average AI Weekly en 2026-09-15T03:54:11+00:00 Occamy-1.0 open 35B edges GPT-5.6 Sol on Claw-Eval average https://aiweekly.co/alerts/zgcm-1-fully-open-7b-claims-to-rival-235b-on-math-search AI Weekly en 2026-09-15T03:54:11+00:00 ZGCM-1: fully open 7B claims to rival 235B on math, search https://aiweekly.co/alerts/vidu-s2-paper-claims-real-time-720p-avatars-and-stream-editing AI Weekly en 2026-09-15T03:54:11+00:00 Vidu S2 paper claims real-time 720p avatars and stream editing https://aiweekly.co/alerts/dream-rsi-paper-reports-162-fewer-agent-calls-vs-simpletes AI Weekly en 2026-09-15T03:30:16+00:00 Dream-RSI Paper Reports 162× Fewer Agent Calls vs SimpleTES https://aiweekly.co/alerts/khan-cites-1934-keppel-case-to-charge-ai-ceos-under-current-law AI Weekly en 2026-09-15T03:30:16+00:00 Khan Cites 1934 Keppel Case to Charge AI CEOs Under Current Law https://aiweekly.co/alerts/omni-streaming-thinking-paper-targets-cross-modal-hallucinations AI Weekly en 2026-09-15T03:30:16+00:00 'Omni-Streaming Thinking' Paper Targets Cross-Modal Hallucinations https://aiweekly.co/alerts/attackers-consumed-600k-in-ai-credits-via-stolen-metr-key AI Weekly en 2026-09-15T03:12:01+00:00 Attackers Consumed $600K in AI Credits Via Stolen METR Key https://aiweekly.co/alerts/ai-chiefs-altman-amodei-musk-agree-to-pace-the-frontier AI Weekly en 2026-09-15T03:12:01+00:00 AI chiefs Altman, Amodei, Musk agree to 'pace the frontier' https://aiweekly.co/alerts/researchers-urge-ai-field-to-drop-agi-as-its-north-star-goal AI Weekly en 2026-09-15T03:12:01+00:00 Researchers urge AI field to drop AGI as its north-star goal https://aiweekly.co/alerts/gurman-maps-apple-h1-2027-iphone-18e-macbook-neo-2-m6-air AI Weekly en 2026-09-15T00:33:16+00:00 Gurman Maps Apple H1 2027: iPhone 18e, MacBook Neo 2, M6 Air https://aiweekly.co/alerts/sakana-ais-pc-alm-trains-1000-layer-nets-without-backprop AI Weekly en 2026-09-15T00:12:01+00:00 Sakana AI's PC-ALM trains 1000-layer nets without backprop https://aiweekly.co/alerts/xai-opens-grok-bot-to-enterprises-with-new-audit-controls AI Weekly en 2026-09-15T00:12:01+00:00 xAI opens Grok Bot to enterprises with new audit controls